In the published article, we presented high-resolution polarimetry data obtained by using JCMT SCUBA-2/POL-2, and compared them with the larger-scale magnetic structure observed by Planck (Planck Collaboration et al. 2020). There was a miscalculation in the analysis of the Planck data for comparison, and the mean position angle of the Planck magnetic field should be corrected from -40° ± 7° to -48° ± 6°. Thus, we replace the descriptions in the published article as follows. In Section 4.2, paragraph 3, the second sentence should read: "The differences between the orientations of IRAS 4A, IRAS 4B, and IRAS 2A are not statistically significant, but they do show significantly different orientations from those of the global B-field observed by Planck (-48° ± 6°; Section 4.1)." In Section 4.1, paragraph 3, the second line should read: "The Planck B-field orientation shows a smoothly and slowly varying field distribution with a position angle of -48° ± 6° in our observed NGC 1333 area." Accordingly, we replace Figures 4, 5, 9, 10, and 17 to reflect the correct Planck data. As described above, this error of modest magnitude is related only to our derivation of the Planck polarization angle, and the JCMT observation results are unaffected. Therefore, all conclusions drawn in the published article are unchanged even after the above correction is applied. The 1 pc scale magnetic field observed by Planck shows a smooth distribution, and the interstellar magnetic field in molecular clouds increases the complexity significantly on the scale of less than 1 pc.